Quoting: Capitalfail67
his data basically says Trouba is barely a good defender but has okay offensive upside... Kuzy has amazing offensive upside but terrible defense.... so if you want a poor mans Carlson who makes the same amount of money you could get trouba but youd regret it bc hes not very good at defense and above average at offense lol. i will take my chances with Kuzy poor defense than trade him for an overpaid average defensman lol SORRY
The data posted actually shows that Trouba isn't very good at generating offense (expected even strength offence value of -0.1, but decent on the PP 1.6), while his defensive contribution (even strength defense 0.6 but -2.2 on the PK) is reasonable. Most of his negative value comes from his use on the penalty kill, which could likely be reduced if this weakness was accounted for. If used properly he would likely add more than 0.1 WAR, but he still isn't all that good.
On Kuznetsov, all value he adds offensively is negated by how bad he is on D, and by the deficit he puts his team at by taking penalties. Additionally, all of his line-mates are made worse by playing with him. This is all happening despite Kuznetsov getting some of the easiest deployment on the team.
Both are overrated, and both aren't very good. It just seems that Trouba is likely less bad than Kuznetsov.
You're free to disagree with the data if you find hockey more enjoyable that way, and though it's true not everything that happens can be efficiently quantified, it's equally as naive to think that someone can watch a given game and lift certainties from from their impressions. The trouble I always have, is if Kuznetsov is such a star-player, why doesn't it show up in his impact? Is he really doing so much that goes undetected that his on-ice metrics are invalid? Seems unlikely. We should use all the tools available to evaluate players so invalidating one method, be it analytics or anecdotal impressions, isn't smart. That being said, people have bias and we miss things all the time, so it's important to be able to cross-reference personal feelings with actual metrics.